MM vs MI vs MC

[Bob from Florida]

Here are similar plots of the AT440MLa - the response reflects a high quality tapered aluminium cantilever - as still used in a number of models currently - I would expect those models to have a similar response - note that on these I have shown both the modelled response as well as the measured response - please ignore the modelled one, as I was experimenting with it at the time (it makes them harder to read, but I figured people would find it interesting nevertheless) - the AT VM cartridges seem to do best with very low capacitance (yes most people won't be able to achieve 60pf...) - and very low C with low R seems the best option.... circa 60pf/30k - at a guess, this would probably be where the VM540ML would be optimal too - as the C rises, the high end rolls off, but the electrical resonance also rises, and can compensate for the rolloff so at 100pf you might want 35k, and at 200pf 47k might be optimal - but you will be accepting a high end boost in exchange for extension - and might prefer to keep things neutral as far as possible and accept a high end rolloff... in which case these AT cartridges seem best with low R loadings eg: 300pf/21k, :

(edit: the smooth lines are modelled, the wavy ones are measured...)

View attachment 201943

View attachment 201944

View attachment 201945

View attachment 201946

Perhaps you are making the case for cartridges unaffected by capacitance. Either moving coil or certain moving iron with low enough inductance to be unaffected.

 

[dlaloum]

Perhaps you are making the case for cartridges unaffected by capacitance. Either moving coil or certain moving iron with low enough inductance to be unaffected.

Not really - the types of cartridges that are unaffected by capacitance, tend to expose the cantilever resonance quite harshly...

If they have a very very good cantilever with very low effective tip mass, the results can be superb (Dynavector Karat!) - but in most cases, even very good cantilevers have too much mass to achieve this, and will have resonances between 10kHz and 20kHz - resulting in a rising high end.

Without the ability to EQ through loading, these resonances become audible (as boosts in the relevant frequencies).

Of course, nowadays, one could choose digital EQ as a valid alternative - but personally I prefer to get the analogue phono out as close to neutral as possible before then applying any digital EQ needed.

 

[Bob from Florida]

Not really - the types of cartridges that are unaffected by capacitance, tend to expose the cantilever resonance quite harshly...

If they have a very very good cantilever with very low effective tip mass, the results can be superb (Dynavector Karat!) - but in most cases, even very good cantilevers have too much mass to achieve this, and will have resonances between 10kHz and 20kHz - resulting in a rising high end.

Without the ability to EQ through loading, these resonances become audible (as boosts in the relevant frequencies).

Of course, nowadays, one could choose digital EQ as a valid alternative - but personally I prefer to get the analogue phono out as close to neutral as possible before then applying any digital EQ needed.

So, I have a Clearaudio Maestro 2 that came with the turntable - recommended load is 100pf but the cable for the Satisfy tonearm is already 174pf before the phone stage adds its 100pf for a total of 274pf. That combination always sounded "bright" to me. I put the Hana SL on with 60 db and 1K ohm load and the brightness is gone and I can listen for hours with no fatigue. The Hana has an aluminum cantilever with Shibata diamond while the Maestro has a boron cantilever with line contact diamond. The Maestro has one of those cantilevers that hang way out in front of the body versus the Hana's under the body. I suspect it is a combination of factors - not just MC versus MM - that makes the Maestro "bright" and the SL very listenable.

On the subject of digital EQ - because I have both analog and digital I would need to employ adjustments to both if I were to get that "picky". Not ready to do that yet, more room treatments first.

 

[dlaloum]

So, I have a Clearaudio Maestro 2 that came with the turntable - recommended load is 100pf but the cable for the Satisfy tonearm is already 174pf before the phone stage adds its 100pf for a total of 274pf. That combination always sounded "bright" to me. I put the Hana SL on with 60 db and 1K ohm load and the brightness is gone and I can listen for hours with no fatigue. The Hana has an aluminum cantilever with Shibata diamond while the Maestro has a boron cantilever with line contact diamond. The Maestro has one of those cantilevers that hang way out in front of the body versus the Hana's under the body. I suspect it is a combination of factors - not just MC versus MM - that makes the Maestro "bright" and the SL very listenable.

On the subject of digital EQ - because I have both analog and digital I would need to employ adjustments to both if I were to get that "picky". Not ready to do that yet, more room treatments first.

Take a look at the measurements for the AT440MLa - it is a very similar design (generator) - albeit with a better cantilever. (Maestro is 420mH and AT440 is 490mH - so in the same range roughly.

You are running (I assume) with 47k resistive load - and 174pf.... and yes most tonearms/setups made in the last 20 years have way too much capacitance for MM setups (!).... as an aside, the cheap/nasty looking cables that TT's from the 70's and 80's came with, are often much better, with around 100pf !!

At 174pf and 47k (and 420mH) you are electrically generating a slight resonance take a look at the smooth green line on the 296pf set of AT440 measurements - the smooth line is modelling the electrical response, where the wavy line is the actual measured response (sum of mechanical and electrical) then look at the same line on the 60pf chart... you will be between these two. - with I expect a slight boost - which you are hearing.

If you drop the R loading to something like 35k you will control/reduce that electrical resonance - and most likely get a response more like what you are seeking.

A Clearaudio Maestro with a boron cantilever should behave very much like an AT150mlx when properly loaded... (the AT150 is 350mH so the electrical resonances are a little different... but not so much that the principles are not applicable!) - they are very very similar generators, and almost identical needles/cantilevers (the Clearaudio uses firmer suspension rubber to give it a lower compliance - and make it more suitable for heavier arms).

Typically for MC's - the cantilever response is exposed due to there being very little EQ adjustment via loading - aluminium cantilevers (like the Hana SL) will tend to behave much like the AT440 with 60pf/83k - you can see there on the modelled response that there is very little if any EQ happening in the audible range.... so the frequency response will be that of the cantilever - and due to its own resonance, that will show up as a high end rise...

Some of the best aluminium cantilevers will have their resonance around 19kHz... they will be flat to beyond 12kHz before starting to rise - which means they can sound very neutral. (Stanton 881)

Most will have their resonances lower down, 8kHz (for the "porkers") up to 14kHz (very good) - if you look at the difference between the modelled lines on those charts, and the measured lines - you can see the AT440's resonance somewhere around 12 to 13kHz - which then requires some EQ to achieve neutrality... The AT440MLa uses a Tapered Aluminium cantilever.

 

[Bob from Florida]

Take a look at the measurements for the AT440MLa - it is a very similar design (generator) - albeit with a better cantilever. (Maestro is 420mH and AT440 is 490mH - so in the same range roughly.

You are running (I assume) with 47k resistive load - and 174pf.... and yes most tonearms/setups made in the last 20 years have way too much capacitance for MM setups (!).... as an aside, the cheap/nasty looking cables that TT's from the 70's and 80's came with, are often much better, with around 100pf !!

At 174pf and 47k (and 420mH) you are electrically generating a slight resonance take a look at the smooth green line on the 296pf set of AT440 measurements - the smooth line is modelling the electrical response, where the wavy line is the actual measured response (sum of mechanical and electrical) then look at the same line on the 60pf chart... you will be between these two. - with I expect a slight boost - which you are hearing.

If you drop the R loading to something like 35k you will control/reduce that electrical resonance - and most likely get a response more like what you are seeking.

A Clearaudio Maestro with a boron cantilever should behave very much like an AT150mlx when properly loaded... (the AT150 is 350mH so the electrical resonances are a little different... but not so much that the principles are not applicable!) - they are very very similar generators, and almost identical needles/cantilevers (the Clearaudio uses firmer suspension rubber to give it a lower compliance - and make it more suitable for heavier arms).

Typically for MC's - the cantilever response is exposed due to there being very little EQ adjustment via loading - aluminium cantilevers (like the Hana SL) will tend to behave much like the AT440 with 60pf/83k - you can see there on the modelled response that there is very little if any EQ happening in the audible range.... so the frequency response will be that of the cantilever - and due to its own resonance, that will show up as a high end rise...

Some of the best aluminium cantilevers will have their resonance around 19kHz... they will be flat to beyond 12kHz before starting to rise - which means they can sound very neutral. (Stanton 881)

Most will have their resonances lower down, 8kHz (for the "porkers") up to 14kHz (very good) - if you look at the difference between the modelled lines on those charts, and the measured lines - you can see the AT440's resonance somewhere around 12 to 13kHz - which then requires some EQ to achieve neutrality... The AT440MLa uses a Tapered Aluminium cantilever.

This is the Maestro calculated resonance peak from alignmentprotractor.com
First time I plugged the numbers in and saw his result I understood why the Maestro sounded the way it did.

 

[dlaloum]

This is the Maestro calculated resonance peak from alignmentprotractor.com
First time I plugged the numbers in and saw his result I understood why the Maestro sounded the way it did.

View attachment 202420

Keep in mind that this is the calculated electrical resonance - and not the cantilever resonance/performance

You need to sum the two to get the actual performance of the cartridge...

Typically you would measure a cartridge/stylus performance - deduct the calculated electrical resonance - that then gives you the cantilever/stylus performance... once you have that, you can then calculate other electrical resonance profiles, and estimate what the end result would be by adding back your cartridge/stylus performance.

It is not totally perfect, as it does not take into account certain losses in the system eg: eddy current losses in the magnetic core - which can be minimised by designs that use a laminated core... eg: the TOTL Audio Technica VM's (AT150, AT540), Shure V15V's and others... the more prosaic models eg AT95/VM95, Shure M97 typically use non laminated cores. (which I am pretty sure also means that the Clearaudio's don't have a laminated core - as the 95 series doesn't feature lamination)

 

[Bob from Florida]

Keep in mind that this is the calculated electrical resonance - and not the cantilever resonance/performance

You need to sum the two to get the actual performance of the cartridge...

Typically you would measure a cartridge/stylus performance - deduct the calculated electrical resonance - that then gives you the cantilever/stylus performance... once you have that, you can then calculate other electrical resonance profiles, and estimate what the end result would be by adding back your cartridge/stylus performance.

It is not totally perfect, as it does not take into account certain losses in the system eg: eddy current losses in the magnetic core - which can be minimised by designs that use a laminated core... eg: the TOTL Audio Technica VM's (AT150, AT540), Shure V15V's and others... the more prosaic models eg AT95/VM95, Shure M97 typically use non laminated cores. (which I am pretty sure also means that the Clearaudio's don't have a laminated core - as the 95 series doesn't feature lamination)

Did not think about eddy currents in a cartridge! Although the coil wire is already so thin I would think the influence of the eddies in the core material would be pretty small on the wire. However, I also can see going with laminations in the core would help with field homogeneity by minimizing the eddies.
I once bent an aluminum cantilever on an expensive moving coil cartridge. After I calmed down enough to consider my options, I spent the next 1.5 hours tediously manipulating that cantilever until it was straight again. Straight with a minor dimple in the aluminum pipe. It always sounded better to me afterwords......

 

[Digital1955]

Straight with a minor dimple in the aluminum pipe. It always sounded better to me afterwords......

You work hardened the aluminum. Sounds like you should offer it as a service

 

[eliash]

This is the Maestro calculated resonance peak from alignmentprotractor.com
First time I plugged the numbers in and saw his result I understood why the Maestro sounded the way it did.

View attachment 202420

This is the Maestro calculated resonance peak from alignmentprotractor.com
First time I plugged the numbers in and saw his result I understood why the Maestro sounded the way it did.

View attachment 202420

Btw., did you take the winding capacitance into account?
...some months ago, I did the same calculations for the Nagaoka MP-500, which has a very high inductance of about 800mH, therefore requiring very little specified cap loading of about 150pf. Calculation showed no resonance and only a gentle decline above 10KHz. When taking the winding capacitance into account, assuming 180pf (higher proposed circuit model value, because of the high inductance w. many windings) summing up to app. 300pF total load results in a nice el. resonance with significant decline above 10KHz and therefore better compensating a distinct boron cantilever resonance (as seen from MC/boron carts' FRs)...
In your case, because of the lower inductance, one could assume maybe some 120pf (lower proposed circuit value) of add'l winding capacitance to hit the truth in el. behaviour...

 

[Bob from Florida]

You work hardened the aluminum. Sounds like you should offer it as a service

I would not recommend that procedure. I was simply "lucky".

 

[Bob from Florida]

Btw., did you take the winding capacitance into account?
...some months ago, I did the same calculations for the Nagaoka MP-500, which has a very high inductance of about 800mH, therefore requiring very little specified cap loading of about 150pf. Calculation showed no resonance and only a gentle decline above 10KHz. When taking the winding capacitance into account, assuming 180pf (higher proposed circuit model value, because of the high inductance w. many windings) summing up to app. 300pF total load results in a nice el. resonance with significant decline above 10KHz and therefore better compensating a distinct boron cantilever resonance (as seen from MC/boron carts' FRs)...
In your case, because of the lower inductance, one could assume maybe some 120pf (lower proposed circuit value) of add'l winding capacitance to hit the truth in el. behaviour...

Curious as to the source of the inductance number. I looked for that type of data for the Nagaoka and found none.

 

[mackat]

Could someone please point me to what is necessary to create the frequency graphs that often appear in this thread? I believe I saw some portions of the info in different posts, but was unable to piece together exactly what needs to be installed and how it works.

 

[Bob from Florida]

You work hardened the aluminum. Sounds like you should offer it as a service

Sample picture of "Aluminum Cantilever" hardening service. Look closely and the dimple is visible.

 

[eliash]

Curious as to the source of the inductance number. I looked for that type of data for the Nagaoka and found none.

Curious as to the source of the inductance number. I looked for that type of data for the Nagaoka and found none.

Neither did I, so I measured with an inductance meter (as posted on p.22 of this thread), right 790mH (590Ohms), left 830mH (605Ohms), suggesting a target of around 800mH (geometric mean). At first, I thought the coil resistance would disturb this measurement, but checking the instrument later on showed that it suppresses the real part (2%+5digits accuracy, 325Hz measurement frequency) quite well. (Btw.: exact mathematical accordance between resistance and inductance deviations of both channels!)

Due to the low measuring frequency, the coil winding cap doesn't have any influence on the result.
In order to extend the high frequency range as much as possible and limit <10KHz overshoot, (in my setup) I implemented the lowest attainable cap load of the el. chain.
All RFI caps removed from the preamp module and preamp chassis resulting in little less than 45pf load (the old-design, but very low noise opamp AD745 with huge input JFETs already contributes 20pF), short RG-180-based 50pF phono-cable, plus 15pF tonearm-cabling, totaling in 110pF. All values verified & measured. Assuming add'l 180pF as winding cap, the result from your quoted calculator is this:

One should note the nice low overshoot of 0.5dB around 6KHz and the attenuation of -3 to -5dB in the range where the cantilever resonance should be present. Unfortunately I don't have the widely used CBS test record at hand here, but direct-switch-over listening tests yield good correspondence between equally leveled and similarly mastered digital and analog program material...

 

[Bob from Florida]

Neither did I, so I measured with an inductance meter (as posted on p.22 of this thread), right 790mH (590Ohms), left 830mH (605Ohms), suggesting a target of around 800mH (geometric mean). At first, I thought the coil resistance would disturb this measurement, but checking the instrument later on showed that it suppresses the real part (2%+5digits accuracy, 325Hz measurement frequency) quite well. (Btw.: exact mathematical accordance between resistance and inductance deviations of both channels!)

Due to the low measuring frequency, the coil winding cap doesn't have any influence on the result.
In order to extend the high frequency range as much as possible and limit <10KHz overshoot, (in my setup) I implemented the lowest attainable cap load of the el. chain.
All RFI caps removed from the preamp module and preamp chassis resulting in little less than 45pf load (the old-design, but very low noise opamp AD745 with huge input JFETs already contributes 20pF), short RG-180-based 50pF phono-cable, plus 15pF tonearm-cabling, totaling in 110pF. All values verified & measured. Assuming add'l 180pF as winding cap, the result from your quoted calculator is this:

View attachment 202725
One should note the nice low overshoot of 0.5dB around 6KHz and the attenuation of -3 to -5dB in the range where the cantilever resonance should be present. Unfortunately I don't have the widely used CBS test record at hand here, but direct-switch-over listening tests yield good correspondence between equally leveled and similarly mastered digital and analog program material...

That graph's calculated response looks preferable to the Maestro. I tried the entry Nagaoka back when I had my Merrill Heirloom and it was very listenable.
I did miss your post on page 22 - but freely admit to not reading anywhere close to all of this thread.

 

[JP]

All values verified & measured.

End-to-end active?

 

[USER]

Could someone please point me to what is necessary to create the frequency graphs that often appear in this thread? I believe I saw some portions of the info in different posts, but was unable to piece together exactly what needs to be installed and how it works.

It all depends on your set-up (e.g. mac or pc), but all pertinent info is found in the first 20 or so pages of this thread. But quickly you will need a test record, python program (no real coding experience is necessary, but you are going to have to be somewhat comfortable with coding interfaces--I was able to troubleshoot some small PC compatibility issues with the script and my only experience with "code" is with HTML/CSS and basic MS-DOS command line prompts), the python script (on this thread), recording software (audacity is recommended), and an anti-riaa filter (one that functions in audacity for recordings at 96kHz is also found on this thread). If you have specific questions once you are up and running feel free to ask, but in the meantime take the time to read the thread carefully. It's all there.

 

[eliash]

End-to-end active?

Both,
- preamp: active @-3dB freq. via 1MOhm feed and oscilloscope measurement on inverting input of opamp (>no probe loading of non-inverting input, w/o clipping!),
- cable and tonearm: passive w cap meter (incl. some plausibility checks - e. g. single ch., 2 ch. in parallel...proving low cap readings...)

 

[mackat]

It all depends on your set-up (e.g. mac or pc), but all pertinent info is found in the first 20 or so pages of this thread. But quickly you will need a test record, python program (no real coding experience is necessary, but you are going to have to be somewhat comfortable with coding interfaces--I was able to troubleshoot some small PC compatibility issues with the script and my only experience with "code" is with HTML/CSS and basic MS-DOS command line prompts), the python script (on this thread), recording software (audacity is recommended), and an anti-riaa filter (one that functions in audacity for recordings at 96kHz is also found on this thread). If you have specific questions once you are up and running feel free to ask, but in the meantime take the time to read the thread carefully. It's all there.

Thank you, I appreciate the response. I scanned the thread and got all of the pieces installed. Unfortunately it seems to be giving me an error; this is what I get when I try to run a test (asterisks added for privacy):

cd '/Users/b***e/' && '/usr/local/bin/python3' '/Users/b***e/Untitled.py' && echo Exit status: $? && exit 1


b***e@b***s-air ~ % cd '/Users/b***e/' && '/usr/local/bin/python3' '/Users/b***e/Untitled.py' && echo Exit status: $? && exit 1


Exit status: 0





Saving session...


...copying shared history...


...saving history...truncating history files...


...completed.


Deleting expired sessions...none found.

 

[mackat]

I finally got the Python script to work! Here are the 2 that I've done so far. Definitely not perfect, but fun nonetheless:

V-15 V-MR connected to a Signet MK12T SUT in pass-thru mode into a MoFi StudioPhono into a very old Griffin iMic USB interface, upsampled from 44100 to 96000, equalized to -30 dB, Nyquist reverse RIAA filter applied. Capacitance estimated at ~300 pf.

150MLX into an ART USB Phono Plus, USB powered by my MacBook Air. Recorded at 96000, equalized to -30 dB, Nyquist reverse RIAA filter applied. Capacitance estimated at ~200 pf, but likely more as shown by the rising freq response?